Justin Johnson
Justin Johnson
I haven't tried the docker install, but I'll give it a try and report back.
@mingfeima I totally agree that the CPU benchmark is pretty far from optimal. I don't have a lot of experience with squeezing every last bit of performance on CPU, so...
@renato2099 They are are single-machine runs; however that single machine is one of the machines in our research cluster.
From the README: > All benchmarks were run in Torch. The GTX 1080 and Maxwell Titan X benchmarks were run on a machine with dual Intel Xeon E5-2630 v3 processors...
The CPU machine was the same as the 1080 and Maxwell Titan X machines, so it had dual E5-2630 v3 and 64 GB RAM. Also you really shouldn't take the...
CaffeNet would be pretty easy to add since it can be loaded into torch with loadcaffe. GoogLeNet v1 is doable using [soumith/inception.torch](https://github.com/soumith/inception.torch) and v3 from [Moodstocks/inception-v3.torch](https://github.com/Moodstocks/inception-v3.torch); I'm not aware of...
@vimalthilak GoogleNet v1 is easy, but I ran into a snag with v3; they don't provide dumped weights and their weight-dumping script depends on TensorFlow; I opened an issue but...
@vimalthilak I've added Inception-v1 results.
My speeds are forward / backward times for an entire minibatch of 16 images; they divide by the minibatch size to try and compute a per-image time. You need to...
There is something wrong with these numbers - the 980 Ti should be roughly the same speed as a Maxwell Titan X, and should thus be quite a bit slower...